Information Space Receding Horizon Control for Multisensor Tasking Problems.

Area of Science:

• Robotics and Control Systems
• Operations Research
• Aerospace Engineering

Background:

• Optimal sensor scheduling is crucial for effective monitoring of dynamic targets.
• The space situational awareness (SSA) problem necessitates efficient multi-sensor coordination.
• Existing dynamic programming solutions for multiagent Markov decision processes are computationally intractable for complex scenarios.

Purpose of the Study:

• To develop a computationally tractable solution for optimal multi-sensor scheduling.
• To address the challenges of monitoring dynamical targets in space situational awareness.
• To present a receding horizon approach combined with stochastic optimization for sensor scheduling.

Main Methods:

• A simulation-based stochastic optimization technique is employed for variance reduction and distributed solutions.
• The core problem is framed as a multiagent Markov decision process on the information space.
• A receding horizon strategy is integrated to solve control problems online, avoiding intractable dynamic programming.

Main Results:

• The proposed technique successfully addresses the optimal sensor scheduling problem.
• The receding horizon approach combined with stochastic optimization yields a computationally tractable solution.
• The method was validated on a moderate-scale SSA example that is intractable for existing techniques.

Conclusions:

• The developed receding horizon solution offers a computationally feasible method for optimal multi-sensor scheduling.
• This approach significantly advances the ability to manage sensor resources for space situational awareness.
• The technique provides a practical solution for complex, large-scale multiagent monitoring problems.